Colorectal cancer (CRC) incidence and death rates are increasing among people under age 50.^1,2 In the United States (U.S.), from 2017 through 2021, the incidence of CRC for people under age 50 years was 8.8 cases per 100,000, with the corresponding death rate as 1.8 cases per 100,000 over the 2018–2022 time period.^1,2 In most states, incidence rates are rising among people under age 50 years.¹ Patients under age 50 years are more likely to be diagnosed at a later stage of disease and have high grade histology compared to patients aged 50 and older, and they often receive their diagnosis after having symptoms, rather than through routine screening.^3–5 Studies have found that primary care visits for potential CRC symptoms are common among patients with early-onset CRC in the year prior to their diagnosis.^6–8

In the United States in 2022, 66% of all early-onset CRC patients aged < 45 were diagnosed as late-stage.⁹ Although overall survival is comparable between early-onset and late-onset CRC patients, late and long term effects of treatment can be particularly challenging for younger cancer survivors who may also be planning or raising families, managing the financial burden of cancer, and building careers.^10–13 In 2021, these surveillance trends over the past decade prompted the United States Preventive Services Task Force (USPSTF) to lower the recommended screening initiation age to 45 years for people at average risk for CRC.¹⁴ Around 16% of early-onset CRC patients have a genetic variant that increases risk for cancer.¹⁵

CRC can present with noticeable symptoms such as changes in bowel habits, symptoms related to anemia such as fatigue or shortness of breath during physical activity, unexplained weight loss, abdominal pain, or bloody stools.^6,8,16 Although many of these symptoms may be attributed to other causes and are not specific to CRC, some are deemed “red flag” symptoms and may need timely diagnostic referral for colonoscopy.^8,17 Additionally, having more than one red flag symptom increases the odds of a CRC diagnosis in people under age 50.⁸ Although a number of studies have described CRC symptoms and identified treatment delays, there is still a recognized need for enhanced provider recognition of symptoms and clear follow up and referral pathways, along with on-time screening of patients at average CRC risk starting at age 45 years.^17–19 Since 2004, early-onset CRC deaths rates have continued to rise in the United States, and symptom recognition and timely diagnostic follow-up are important for average-risk patients under age 45 who do not meet the criteria for routine screening.^17,19,20

To date, few studies are available on primary care providers’ (PCPs’) knowledge of CRC signs and symptoms in younger patients and their approaches to diagnostic workup and referral in the United States health care system. The objective of this study was to assess PCP knowledge of CRC signs and symptoms and their diagnostic practices, which may help inform future provider-related interventions by identifying areas to strengthen clinical support. This study builds on prior work on awareness of CRC red flag symptoms by providing a contemporary analysis following the release of the 2021 USPSTF screening recommendations that lowered the starting age for CRC screening for people at average risk for CRC and raised awareness of early-onset CRC.^14,21

In this large, nationally representative panel of PCPs practicing in the U.S., PCPs could identify a median of 4 signs/symptoms from a predefined list of 6 CRC-related signs/symptoms. Awareness was higher for more colon or rectal-specific symptoms such as changes in bowel habits or blood in the stool rather than symptoms reflective of many other conditions, such as abdominal pain. PCPs with more experience and patient load were more likely to recognize symptoms, along with those specializing in internal medicine or family medicine.

In this study, 45% of PCPs would refer patients to diagnostic colonoscopy to further assess rectal bleeding or bloody stools, but only 24% would refer patients presenting with iron deficiency anemia to diagnostic colonoscopy. These referral rates are lower compared to those found in a Colorado study reporting a majority of PCPs (72%) would refer patients to diagnostic colonoscopy for iron deficiency anemia, with slightly over half doing the same for hematochezia (i.e. bright red blood in the stool).²¹ In another study using clinical vignettes of patients with iron deficiency anemia, bidirectional endoscopy to identify intestinal bleeding was underused, and while 77% of survey respondents would follow up with endoscopy for a 35 year old man, less than 20% would for a 35-year old woman with iron deficiency anemia.²² Patients experiencing changes in bowel habits (67% referral rate) or abdominal pain (61% referral rate) were less frequently referred to diagnostic colonoscopy compared to patients with unexplained weight loss (84% referral rate) in another study, although the vast majority of all patients received some type of CRC workup (77%).²³

For symptoms that often occur in CRC, PCPs in this study indicated they typically follow up with lab or physical examinations or assess for family history of CRC or related cancers, and less than half would do a GI referral in all scenarios. Referral practices also varied for the 3 signs/symptoms assessed. Nearly 30% of PCPs indicated they would advise patients experiencing changes in bowel habits to monitor their symptoms. For patients experiencing these symptoms, fewer PCPs (62%) reported they would assess patients for family history of cancer compared to patients experiencing other symptoms, although most would complete a physical workup. This may be a missed opportunity to identify patients at increased risk for CRC based on family history of cancer, because genetic variants that increase CRC risk are more common among people diagnosed with early-onset CRC compared to older patients, and having a family history of CRC is an important risk factor.^24–26

Studies have found that nearly half of CRC patients initially present with blood in the stool or abdominal pain, and around 27% present with altered bowel habits.¹⁶ Given that CRC patients aged < 50 are more likely to present at later stage at diagnosis, some major medical organizations recommend diagnostic colonoscopy for patients presenting with lower gastrointestinal tract bleeding or iron deficiency anemia.^27,28 The current study’s findings suggest that PCPs may prefer to take a more measured approach when presented with patients with CRC red flag symptoms, such as conducting a workup and assessing family history of cancer before considering a diagnostic colonoscopy or GI referral based on a single sign or symptom. These findings align with a study that examined follow up testing after the appearance of symptoms using claims data and found that other tests, such as FOBT/FIT or cross-sectional imaging, were often conducted first before diagnostic colonoscopy.²⁹ This approach may reflect a desire to rule-out other conditions first before following up with GI or colonoscopy referral. However, we found a minority of PCPs would follow up with a stool-based test. Given that we focused on patients under age 50, some PCPs may have a preference to offer a noninvasive test to unscreened patients aged 45 and older. Although this study did not assess colonoscopy availability, limited accessibility to colonoscopy may continue to exist in some rural geographic areas of the United States.^30,31 Therefore, some PCPs may prefer to make colonoscopy referrals as a final step.

Understanding the circumstances that lead to use of stool-based tests for triage or diagnostic purposes may inform future interventions on appropriate clinical follow up for patients presenting with potential CRC-related symptoms. Currently, stool-based tests are recommended by the USPSTF and other major medical organizations only for screening of asymptomatic average-risk patients.^5,14,32,33 In one study, 22% of patients receiving a multi-target stool DNA test were inappropriately tested because they had one or more contraindications for this test, including having a colonoscopy within the past 10 years, a significant family history of CRC, or potential CRC-related symptoms.³⁴

There is evidence that quantitative FIT calibrated to a lower detection threshold may help triage patients needing diagnostic colonoscopy by outperforming referrals based on symptom assessment alone.³⁵ The Association of Coloproctology of Great Britain and Ireland (ACPGBI) and the British Society of Gastroenterology (BSG) have produced joint guidance on management of patients with CRC symptoms by incorporating quantitative FIT results as a triage tool for urgent diagnostic colonoscopy.^36,37 Another initiative is the use of timed diagnostic pathways (i.e. Faster Diagnosis Standard) to ensure patients are not lost to follow up and receive either a diagnosis or followed until resolution of symptoms.³⁸ This approach could be helpful in U.S. settings with fewer resources and isolated rural areas where accessing diagnostic colonoscopy may be more challenging and adding additional strain on endoscopists may increase wait times for other patients needing immediate diagnostic colonoscopy. The current study’s findings may also be influenced by the increasing use of FIT over the past several years rather than colonoscopy for routine screening and patient preferences for noninvasive stool tests.^39–42 Additionally, trends in colorectal cancer screening prevalence show that recent increases in screening may have been driven by use of sDNA/FIT, and although most PCPs and specialists still have a preference for colonoscopy, many recognize that stool-based tests are ideal for patients reluctant to undergo colonoscopy as a screening test.^43,44

Clinical decision support and refined risk assessment tools may help clinicians decide on appropriate diagnostic colonoscopy referral or the most appropriate personalized CRC screening initiation age based on polygenic risk scores, personal risk factors, or biomarkers.^45–47 The best approach to addressing early-onset CRC may be offering tailored screening to patients at increased risk based on family history of cancer and risk assessment algorithms, and ensuring patients that do have symptoms are followed up with the appropriate diagnostic testing in a timely manner, while minimizing the number of patients lost to follow-up in the process.¹⁷ Some researchers have proposed frameworks to address early-onset CRC through timely diagnosis by increasing awareness among young adults, proposing symptom triage and risk-adapted diagnostic pathways, increasing the collection of cancer family history, increasing health care provider awareness, and advancing EHR tools, machine-learning algorithms, and artificial intelligence to support health care providers in evidence-based decision-making.^17,19

Some medical specialties, such as OB/GYNs, may benefit from educational outreach on early-onset CRC, given that many OB/GYNs function as PCPs to women of childbearing age. Partnering with medical organizations that represent these clinicians to deliver continuing medical education credits may help to raise awareness on appropriate evidence-based strategies to manage patients with potential symptoms of early-onset CRC.

The current study had several limitations. First, questions were included on a panel survey, and not a population-based sample. The sample tended to be younger than practicing PCPs in the United States, but other demographic characteristics were similar.⁴⁸ Second, the self-reported responses could be influenced by social desirability bias. Third, there was limited survey space to ask in-depth questions using clinical vignettes that included a patient’s detailed history. With this information, it’s possible that patients presenting with a constellation of symptoms or having a family history of cancer may warrant GI referral or diagnostic colonoscopy by PCPs. However, this study had a large sample of PCPs from across the United States, including those practicing in rural areas and in different types of practice settings. Although the current study’s questions were not cognitively tested, subject matter experts in survey design reviewed the questions during the survey development period.

In conclusion, this study found that awareness of most CRC-related signs and symptoms was fairly high among PCPs, but opportunities exist to increase sign and symptom awareness among OB/GYNs and Physician Assistants, and PCPs with less experience or patient load. Additionally, PCPs may benefit from additional training or implementation of clinical decision support tools to increase appropriate referral and follow-up when CRC-related symptoms occur. Assessing the current state of awareness for early-onset CRC and evidence-based practices around risk assessment and on-time screening may help to inform cancer control efforts.